Case Study: Downer Group achieves 80% accurate safety control identification with Crayon’s NLP solution

Using Natural Language Processing to improve safety management at Downer Group

Downer Group, a leading infrastructure and facilities company in Australia and New Zealand, faced the immense challenge of analyzing thousands of safety documents and risk controls. This was a critical part of their "Zero Harm" safety program, but the sheer volume of data made it extremely difficult to optimize their safety management system and provide clear, actionable information to frontline workers.

Crayon developed a solution using a natural language processing (NLP) model built within Azure Machine Learning. The model was trained to automatically identify and tag safety controls mentioned in Downer's vast documentation. This tool achieved up to 80% accuracy in just five weeks, saving significant staff time and providing a foundational AI capability. The solution empowered Downer to efficiently harmonize its safety systems and is paving the way for future AI-driven tools to enhance frontline decision-making and safety.

Downer Group

Mathew Hancock

General Manager